LPP EUV sources have been under discussion for some time. As the requirements for, e.g., smaller and smaller integrated circuit critical dimension lithography and the concomitant requirement for shorter and shorter wavelength light sources, in the ranges of tens of tens of nanometers (e.g., 10-30), the need for a workable EUV light source that can also meet all of the requirements for power, repetition rate, dose stability, and the like requirements the actual requirements for an EUV light source, e.g., for use as a lithography light source, are becoming more clear. By way of example, there are some indications of what the power requirements could be. One way to look at this is to compare reported performance of a laser produced plasma (“LPP”) system, e.g. the TRW/CEO, system, incorporating certain lithography parameters that appear to be system requirements, with proposals for a deep plasma focus system, a variety of discharge produced plasma “(DPP”) systems. Reported numbers for the TRW/CEO system are shown below in Table I.
TABLE ITRW/CEO LPPCollected EUV power at intermediate focus100W**(“I.F.”)Collector optical transmission 55%*EUV power into collector181WGeometric collection efficiency5str/2π strEUV power into 2π str227WLaser-to-EUV conversion   1.0%“Pump” power into vessel22,700WElectrical-to-laser conversion 3%Wall plug electrical power756,666W*According to a TRW/CEO poster paper given at the 2003 SPIE.**According to requirements being stated by potential customers for EUV light sources.
While some systems in use, e.g., in an integrated circuit fabrication facility require power in the range of a kilowatt, the likelihood is that there would be required many more scanners using EUV light sources per fab than, e.g., ion implanters or rapid thermal annealing systems, also requiring this type of projected input power. There is a clear need for improvements to proposals for EUV light source efficiencies.
One area of critical importance to the overall efficiency of such an EUV light source is the collector. Many issues of collector efficiency need to be addressed, including debris management, which can interfere with the ability to deliver the required light energy to the intermediate focus and also decrease economic efficiency of the light source if debris, e.g., requires frequent replacement of the collector due to inability to control debris deposition over time. Proposals for a collector system have been discussed in the co-pending application Ser. No. 10/798,740 entitled COLLECTOR FOR EUV LIGHT SOURCE, filed on Mar. 10, 2004, the disclosure of which is hereby incorporated by reference.
With, e.g., a 10% electrical-to-laser conversion efficiency then the required wall plug power becomes 227,000 W. This value is essentially the same as for the discharge produced plasma (“DPP”). If TRW/CEO can also achieve their stated goal of doubling the laser-to-EUV efficiency, then the required wall plug power becomes 113,500 W. Of course, the methods of increasing this conversion efficiency will likely apply to the DPP and thus the DPP wall plug requirements will also drop by half.
One of the driving forces behind the design of an EUV lithography light source and, e.g., the selection of target material, collector strategy, discharge produced plasma (“DPP”, e.g., deep plasma focus (“DPF”) or laser produced plasma (“LPP”) and the like is the requirement by the lithography tool manufacturers regarding the level of out-of-band radiation, e.g., produced by an LPP source, e.g., with a 248 nm drive laser. Since the EUV multi-layer mirrors exhibit high reflectivity to the UV region and many of the proposed EUV photoresists are sensitive to UV/DUV, it is critical that the source does not produce a large amount of radiation, e.g., in the 130-400 nm range. With a 248 nm drive laser, as opposed to an infrared drive laser, even a small amount of scattered laser light may lead to high levels of UV radiation from the EUV source.
The currently contemplated full specification for out-of-band radiation for a production EUV source is listed below in the wavelength ranges of interest and the allowed ratio to the in-band, e.g., at 13.5 nm energy.
Allowed Percentage (relativeRangeto 13.5 nm in-band)10-40nm100%40-130nm100%130-400nm1%400-800nm100%>800nm0.05%
Therefore all radiation, e.g., between 130 nm and 400 nm must be less than 1% of the in-band 13.5 nm radiation. Thus, if one assumes, e.g., a 2% contribution into in-band EUV then one must also have only a 0.02% conversion efficiency into the 130-400 nm band. This is an incredibly tight requirement, for both LPPs and DPPs.
Behavior of expanding laser produced plasma and/or the effects of magnetic fields on plasmas have been modeled and studied, as discussed, e.g., in H. Pant, “Behavior of Expanding Laser Produced Plasma in a Magnetic Field,” Physica Scripta, Vol. T75 (1998), pp. 104-111; Tillmack, Magnetic Confinement of LPP, UCSD Report and Abramova, “Tornado Trap, the disclosures of which are hereby incorporated by reference.